INDEX
Explanations
recipes or food-related instructions
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
453
+0.16
0.5%
50
+0.14
0.5%
752
+0.13
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1804
+0.16
0.08
453
+0.14
0.07
2011
+0.13
0.05
Negative Logits
in
-0.91
a
-0.91
,
-0.90
for
-0.90
so
-0.87
all
-0.85
as
-0.84
even
-0.84
and
-0.83
or
-0.83
POSITIVE LOGITS
alkoh
2.24
utop
2.23
silikon
2.18
solidar
2.10
karton
2.08
kosme
2.06
keramik
2.05
makro
2.04
marte
2.04
meras
2.03
Activations Density 0.408%